Actually, I'm not sure it's important to Sun. To my knowledge the decision to move to apache was made by the Roller community and had nothing to do with Sun.
That being said, I agree with Matt. I am just here to develop code and make the project better. From all the recent discussions it sounds like being part of Apache will force us to backtrack quite a bit for no good reason. So I agree that if we can't find a way to resolve the legal issues fairly easily then maybe we shouldn't be an Apache project yet. -- Allen On Mon, 2005-11-07 at 08:29, Matt Raible wrote: > I know that having Roller as an Apache project is important for Dave > and Sun. However, I think it's important to state my opinion. If > it's going to require a significant re-write of Roller - my vote is to > back out of the incubator and continue as an open-source project on > java.net (or another location that doesn't have LGPL restrictions). > > Matt > > On 11/7/05, Anil Gangolli <[EMAIL PROTECTED]> wrote: > > Inline comments below... > > Henri Yandell wrote: > > > > > > > >For 2.0 I think we should be trying to get a clear LGPL message, or > > >proto-message; and get things dealt with prior to the release. > > > > > >Effectively this means: > > > > > >1) Removing 'bad' jars from SVN. > > >2) Modifying install process/build instructions to instruct user to > > >download the relevant jars themselves. > > > > > > > > For item (2), I'd like to understand the boundary of what we can do. > > For example, can we provide ant scripts / other build infrastructure > > that is generic, and configured specifically to download these jars? If > > so, we should start building a network-based pre-build target and an > > installer in the Roller 2.x line. > > > > > > >3) Ensuring nobody is concerned with the various usages of LGPL'd code. > > > > > >Hen > > > > > > > > For Hibernate, we do import org.hibernate.* classes and use Hibernate > > APIs directly currently. So we are currently in the "worst" category of > > "offenders". > > > > A JSR-220-based rewrite would entail quite a bit of effort for 2.0, and > > I think will force us to require Java 5 / JDK 1.5.x. The main thing > > that bugs me about this direction is the cramped timing with respect to > > the maturity levels of JSR 220 and the Hibernate support for it, > > although alternate supporting implementations might be considered along > > this line as well. > > > > > >
